Introduction to circuit analysis and design

Introduction to Circuit Analysis and Design

Volumn , Issue , 2011, Pages 1-768

  Glisson Jr , Tildon H ^a  

^a North Carolina State University   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 84892328064     PISSN: None     EISSN: None     Source Type: Book    
DOI: 10.1007/978-90-481-9443-8     Document Type: Book

References (211)

1. Current in an arc is visible, but unless we are designing an electric welder, we usually try to avoid arcs.
2. The essential property of time is perhaps best captured by Yogi Berra's definition: "Time is what keeps everything from happening all at once." The essential properties of charge are given in Chapter 2.
3. One subject's laws can be another subject's derived results; for example, Ohm's law (a law in circuits courses) is a derived result in some more fundamental courses.
4. A paradigm example is one that illustrates application of a law, derived result, or procedure in the simplest possible setting, uncluttered by need for other laws, results, or procedures.
5. SI System is redundant, because SI stands for Systeme Internationale, which (obviously) is French for international system. Nonetheless, "the SI system" is what everyone calls it.
6. Actually, which seven dimensions are defined as basic is a matter of choice. Any seven independent dimensions would do, but history and common sense have ruled in favor of the seven given here. There is much wider agreement on which seven quantities are basic than on the standard units for those seven.
7. Of course, the coefficients of a Taylor series can be dimensioned, and one could define different series for (e.g.) the logarithm of different physical quantities. Such an approach would be cumbersome, at best.
8. Some mathematical software (e.g., Mathcad) that allow units to be used insist that they be attached to all quantities, even those having zero magnitude. For example, if you ask Mathcad to compare 10 s to 0, you will get an error. You must compare 10 s to 0 s.
9. The abbreviation lb stands for libra, an ancient Roman unit of weight which was presumably the predecessor to the English pound. See Yunus A. Cengel and Michael Boles, Thermodynamics: An Engineering Approach (3rd Ed.), McGraw-Hill, 1998.
10. An especially complete table of conversion factors is given in the Handbook of Chemistry and Physics (76th Ed.), edited by David R. Lide, CRC Press, New York, 1995.
11. Resistance is defined in Chapter 2.
12. After the French physicist Charles Augustin de Coulomb (1736-1806).
13. In practical applications, two charges can be considered point charges if their radii are much smaller than their separation.
14. Total charge is the algebraic sum of charge. Thus, the net charge of a hydrogen atom, which consists of one electron and one proton, is zero.
15. We have Ben Franklin to thank for this unfortunate accident of history. He decided to call the static charge induced on glass by rubbing the glass with silk positive and stated that the direction of current is from positive to negative (from glass to silk). Had he made either of these choices the other way, the positive direction of current would be in the direction of electron flow.
16. Explicit time-dependence often is omitted for the sake of brevity. Thus, in (2.2) q stands for q(t) and i for i(t). Also, current defined by (2.2) is called conduction current. When you study electromagnetism, you will learn about another kind of current called displacement current, that does not involve flow of charge. Conduction current is the current of interest in ordinary circuit analysis, where we are relatively unconcerned with the inner workings of circuit elements.
17. After the French Mathematician and Physicist André Marie Ampére (1775-1836), who discovered basic laws of electromagnetism.
18. To avoid using vector notation, we assume the force f(x) is directed along the x-axis, in either the same direction as the motion or opposite to the direction of motion.
19. After the English physicist James Prescott Joule (1818-1889).
20. In this discussion, we are concerned only with energy exchanged between the gravitational field and a single mass. There may be other forces acting on the mass in addition to that due to gravity, but they do not affect the energy exchange between the mass and the field; for example, we can throw a mass toward the ground, in which case additional energy is transferred to the mass, but the energy transferred to the mass from the field is unaffected by the throwing.
21. Note that electric potential is not the same as potential energy. The unit of electric potential is that of energy divided by charge.
22. After Count Alessandro Volta (1745-1827), an Italian inventor who discovered hydrolysis and invented the battery and the electric condenser (now called a capacitor).
23. The terms rise and drop arise from analogy with gravity, where a mass gains potential energy if we raise it and loses potential energy if we lower (or drop) it.
24. In general (and usually), both voltage and current vary with time.
25. After the German physicist Georg Simon Ohm (1787-1854).
26. In reference works, resistivity often is specified in μΩ cm (10-8 Ω m).
27. Metals are good conductors because the outer electrons of metal atoms are only loosely held and are relatively free to "drift" through the solid under the influence of an applied voltage.
28. Resistivity can also vary with pressure, strain, strength and direction of an applied magnetic field, and other things. The operation of various transducers (e.g., strain gauges) is based upon such effects.
29. Handbook of Chemistry and Physics (76th Ed.), edited by David R. Lide, CRC Press, 1995.
30. Ibid. The temperature coefficients were obtained from the slopes of linear-least-squares fits to the resistivity data given there.
31. We treat transformers and inductors in Chapter 9.
32. So called because earth ground (the grounded part of an electrical receptacle) is often the reference point.
33. The meaning of independent in this context is described in the sequel.
34. The polarity marks and arrows and the letters n and i (without subscripts) define the voltage and current that appear in the terminal characteristic. They are not part of the symbol for the element.
35. After the German physicist Gustav Robert Kirchhoff (1824-1887).
36. Kirchhoff's current law can also be stated as "the sum of currents entering a volume equals zero" or as "the sum of the currents entering a volume equals the sum of the currents leaving the volume." For the present, it is best to stick with one version of the law.
37. In "leaving" and "entering," we are referring to the direction of the arrow associated with the current.
38. Here, "current" includes both conduction current and displacement current.
39. Kirchhoff's voltage law can also be stated as "the sum of voltage rises around any closed path equals zero" or as "the sum of the rises equals the sum of the drops." Again, for the present, it is best to stick with one version of the law.
40. Review the definitions of independent current and voltage sources in Section 3.6.
41. Loosely, an element or circuit is regarded as a source if the element or circuit delivers energy to a circuit and is regarded as a load if it removes energy from the circuit (converts electrical energy to another form); e.g., a generator is ordinarily regarded as a source and a motor is ordinarily regarded as a load. As another example, from the perspective of an amplifier in a home audio system, a CD player is a source and a loudspeaker is a load.
42. See Section 3.10 in Chapter 3.
43. In Chapter 6, we introduce sources called dependent sources. In general, the look-back method fails for circuits containing such sources
44. A planar circuit is one that can be drawn in two dimensions without crossing conductors.
45. Also, a source can dissipate power during some time intervals and deliver power during others.
46. Using electronic devices, it is possible to construct a two-terminal circuit that has the properties of a negative resistance; however, such a circuit is not a physical resistor.
47. Certain two-terminal circuits are equivalent to a negative resistance at their terminals.
48. The ambient-air power-dissipation rating of a component can be exceeded if heat sinks and/or forced cooling are used.
49. Common ratings for ordinary composition resistors are 1/8, 1/4, 1/2, and 1W. Composition resistors having ratings up to 5Ware easily obtained. If a much higher power-dissipation rating is warranted, one can use metallic-film or wire-wound resistors, or use forced cooling, or both.
50. Many books give guidelines for sizing heat sinks with and without forced-air cooling. See, for example, Paul Horowitz and Winfield Hill, The Art of Electronics (2nd Ed.), Cambridge University Press, pp 312 ff.
51. The power dissipated by a modern high-performance microprocessor is on the order of 100 W cm -2.
52. Equation (5.11) states that the total power dissipated equals zero. It is also true that the total power delivered equals zero, or that the total power dissipated equals the total power delivered (in a circuit).
53. This condition is more stringent than necessary, but in this book, we do not need to compute time averages of functions not satisfying the condition.
54. In this context, mean is a synonym for average.
55. For this reason, the rms amplitude of a current or voltage also is called the effective value of the current or voltage; however, this terminology is falling into disuse.
56. Review Section 3.1 of Chapter 3.
57. Operational amplifiers are treated in Chapter 7.
58. Capacitors and inductors are treated in subsequent chapters
59. In general, the source current or voltage can be a linear or nonlinear function of one or more other currents and voltages. More generally, the input or output can be a function of a nonelectrical quantity, such as temperature, pressure, or torque. For example, an electric motor can be regarded as a controlled source whose input is current and whose output is torque.
60. These names are contractions of transfer resistance and transfer conductance, respectively.
61. Remember that Fig. 6.20 is the structure of a two-port model, not of the associated physical circuit.
62. Two-port models can be extended to cases where a circuit in question contains independent sources, but we have no need of that generalization in this book.
63. That is, neither is necessarily associated with any particular physical resistor.
64. These elements are treated in subsequent chapters.
65. Peak amplitudes can be (and often are) used if the voltages or currents involved have the same waveform
66. In integrated op amps, transistors with the base connected to the collector often serve as diodes.
67. We treat capacitance in Chapter 8.
68. Real op amps have additional terminals whose functions you will learn when you study electronics.
69. Because of the manner in which the internal circuitry is connected to the supplies, the supplies or supply terminals often are called rails.
70. Also, a small dc offset voltage afflicts virtually all op amps, as described in Section 7.8.
71. Although vpm is small, the individual voltages vp, vn are not necessarily small.
72. Again, the gain of an op amp is so large that if the voltage vp - vn is greater than μ0/VCC, the amplifier will saturate.
73. Achieving gains less than 10 with an inverting amplifier is attended by several problems, as discussed in Chapter 17.
74. This parameter is not specified on manufacturers' data sheets and is difficult to measure. The values given here are deduced from other performance data. Related topics are discussed in more detail in Chapter 17.
75. The output stage of an op amp is typically a so called push-pull configuration of two transistors, where one transistor amplifies the positive parts of an applied voltage and the other amplifies the negative parts.
76. This is the power delivered by the supply to the output stage. A typical op amp is a three-stage amplifier, and even if no input is applied (even if i0 1/4 0), some dc power is required to keep the op amp in an active (ready) state.
77. We justify this rule in Chapter 17.
78. Resistors do possess thermal memory because resistance changes with temperature and the temperature of a resistor depends upon past current through the resistor. This particular form of memory has not been found useful in signal-processing circuits.
79. So far as we know, the present output of a circuit (or any other physical system) cannot depend upon future values of an input. If it were otherwise, we could predict the future.
80. Recall that the positive direction for current is opposite to the direction of electron flow.
81. Electrons migrate until the work the battery can do on a charge equals the work necessary to move the charge from the positive (upper) plate through the circuit to the negative (lower) one, or, until Kirchhoff's voltage law is satisfied, whichever perspective you prefer.
82. The charge will gradually leak through the imperfect insulator separating the plates, but for a good insulator, the leakage will occur very slowly.
83. James Clerk Maxwell (1831-1879).
84. Further discussion of displacement current is beyond the scope of this book. Conduction current is the dominant current in elements and circuits considered in this book, except in insulating media inside capacitors, where the current is predominantly displacement current. Even in that case, we are concerned only with the conduction current into and out of the terminals of the capacitor.
85. After the English scientist Michael Faraday (1791-1867).
86. In a parallel-plate capacitor, the plates are of equal area and uniformly separated with their edges aligned.
87. Eq. (8.5) is derived in virtually all university-level introductory physics textbooks.
88. Continuity of the voltage across a capacitor also follows from (8.10).
89. Section 8.9 treats energy stored in the electric field in a capacitor.
90. See Section 8.10.8 for an example.
91. Corcoran, George F., and Henry R. Reed, Introductory Electrical Engineering, Wiley, New York, 1957, pp 419-426.
92. The equations describing the circuit are nonlinear and cannot be solved in closed form. The analysis was performed numerically, using methods described in John R. Hauser, Numerical Methods for Nonlinear Engineering Models, Springer, 2009.
93. All these values were obtained from the numerical analysis.
94. Direct coupling also is called dc coupling and capacitive coupling also is called ac coupling.
95. Photograph courtesy MTM Scientific, Corp.
96. After the German physicistWilhelmEduardWeber (1804-1891).
97. Of course, if you actually do this with a "live" wire, all of your appendages could point in different directions.
98. After the English chemist and physicist Michael Faraday, (1791-1867). Faraday's law of induction is treated in virtually all introductory physics textbooks.
99. This assertion must be modified for flux produced in a magnetic material by relatively large currents. We ignore that refinement here.
100. After the American physicist Joseph Henry (1791-1878).
101. This relation is derived in all introductory, university-level physics textbooks.
102. Grover, F.W., Inductance Calculations, Dover, New York, 2004.
103. For example, if the current through an inductor has a triangular waveform, then the voltage across the inductor has a rectangular (piecewise-constant) waveform.
104. This terminology is borrowed from Paul Horowitz and Winfield Hill, The Art of Electronics, (2nd Ed.), Cambridge University Press, New York, 1980.
105. The ignition coil in an automobile is actually a transformer.
106. In fact, the motor might not even run, unless it is a universal or ac/dc motor
107. This is equivalent to assuming that the permeability of the core material is constant and independent of position.
108. Between-conductor capacitance also contributes to the coupling. Also, because of nonlinearities in power transformers, the third and fifth harmonics (180- and 300-Hz components) also are present in current on power lines. Telephone companies use filters to eliminate (block) components at and below 300 Hz.
109. Some define the turns ratio as N1=N2.
110. In certain configurations, a special kind of transformer, called an autotransformer, can pass dc. See Problems P 9.93 and P 9.94 at the end of the chapter.
111. As a practical matter, there is little reason to do otherwise.
112. Yes, the plural of chassis is chassis: Spelled the same, but pronounced differently.
113. Balun is a contraction of balanced-to-unbalanced.
114. Paul Horowitz and Winfield Hill, The Art of Electronics, Cambridge University Press, New York, 1989, p 304.
115. You will be able to show this after studying Chapter 12.
116. Wheeler, H.A., Simple Inductance Formulas for Radio Coils, Proc. IRE, vol 16, p 1398 (Oct. 1928).
117. Mathematicians, physicists, and others use i to denote the imaginary unit. Electrical engineers use j because they use i to denote electric current.
118. What we call rectangular coordinates also are called cartesian coordinates and the rectangular form of a complex number also is called the cartesian form of the number.
119. Because we use radian measures exclusively, we do not attach the dimensionless unit rad to numerical values for angles.
120. Mathematicians refer to the magnitude as the modulus and to the angle as the argument.
121. Software packages that provide for dimensioned variables require consistent use of units (if used at all), even for zero values. Also, note that we require a unit for zero temperature, because zero Celsius, zero Fahrenheit, and zero Kelvin are all different.
122. Mathematicians call the forced response the particular integral and the unforced response the complementary solution (or function). The names used here are more appropriate (and suggestive) for physical problems.
123. Note that s (italic) is a variable and s (roman) is a unit (second). In written work, you might want to distinguish the unit s from the variable s by underlining the former.
124. The sum of functions having different time dependences cannot be identically zero (cannot equal zero for all time); e.g., a sinusoidal function of time cannot cancel a real exponential function of time.
125. Be aware that s (italic) is a variable and s (roman) is a unit. This usage, which is almost universal, is doubly unfortunate because the unit of s is s. In written work, you should underline the unit s to avoid confusion.
126. A circuit containing only resistors, capacitors, and dependent sources (active devices) can be underdamped, as can a circuit containing only resistors, inductors, and dependent sources.
127. Chapter 16 introduces Fourier analysis. 2Angular frequency is called radian frequency in some books.
128. In Chapter 14, we express angles in degrees, as is conventional in the electric power industry.
129. Determining frequency and initial phase from such a graph requires calculations, albeit simple ones.
130. More generally, within any whole number of periods.
131. If two sinusoids have different frequencies, then one alternately leads and lags the other. The duration of an interval during which one leads (or lags) the other depends upon the frequencies.
132. In this book, a tilde (∼) denotes a complex representation of a real quantity; thus, x̃(t) denotes the complex representation of a real signal x(t).
133. Recall that the positive direction for an angle is counter-clockwise.
134. It is meaningless to add the phasors for sinusoids having different frequencies.
135. Fourier analysis is introduced in Chapter 16.
136. Section 12.16 treats superposition in the context of sinusoidal excitation.
137. Indeed, this is why the cosine, rather than the sine, is chosen as the standard form for a sinusoidal current or voltage.
138. If there is more than one source, we may use superposition and consider one at a time.
139. Of course, we can also use admittances. But conventionally, elements are represented by their impedances in circuit diagrams, even if circuit equations are subsequently written in terms of admittances.
140. The effective resistances of physical inductors and capacitors are not zero and their reactances are not exactly those given in (12.48). These and other departures of physical components from ideal behavior are discussed in Section 12.22.
141. Note that for any particular load at any particular frequency, reactance and susceptance have opposite signs (if non-zero).
142. These remarks pertain to passive circuits.
143. We define the admittance triangle for sake of completeness; however, admittance triangles are rarely seen in practice.
144. Because the standard form for a sinusoid is a cosine, a dc component can be treated as a sinusoid whose frequency is zero.
145. For example, at sufficiently high frequencies, an inductor exhibits shunt capacitance. Section 12.22 is an introductory treatment of such parasitic effects
146. As we show in a subsequent chapter, it is possible to emulate inductance using an active device and capacitance. Thus it is possible for an active circuit containing only active devices, resistors, and capacitors to exhibit resonance. Also, as discussed below, a physical component (such as a capacitor) is not ideal and can exhibit all three of resistance, capacitance and inductance and can be self-resonant at a sufficiently high frequency.
147. Remember that the voltage across an inductor is proportional to both the inductance and the rate of change of the voltage.
148. Actually, the circuit considered here only approximates a true gyrator, but provides a useful approximation at minimal cost.
149. The impedance at resonance is of interest because it determines the maximum current required of a circuit driving the impedance.
150. Adapted from Bogatin, Eric. Signal and Power Integrity - Simplified. Prentice-Hall, Englewood Cliffs, NJ. 2010.
151. Leslie Green, RF Inductor modeling for the 21st Century, EDN, September 27, 2001.
152. See http://www.coilcraft.com/models.cfm
153. See www.capacitorindustries.com
154. http://www.eigroup.org/cmc/downloads/r2-cmc/r2-cmc-v1.0-r0.0-2005nov12. pdf
155. Equation (13.22) is a degenerate case of Tellegen's theorem, which is a remarkable property of any pair of circuits having the same topology. See Artice M. Davis, Linear Circuit Analysis, PWS Pub. Co., (1998), pp 1052ff.
156. The proof can be skipped without ill effect.
157. It is possible for the real part of the impedance of a load containing one or more dependent sources to be negative. The terminology introduced here is used only in references to source-free loads.
158. Again, these remarks pertain to source-free loads. A load containing dependent sources, capacitors, and resistors (no inductors) can be made to appear inductive, as shown in a subsequent chapter.
159. Line voltage varies somewhat with load because of losses in transmission, but not nearly to the degree that line current varies with load.
160. Superposition of real power is treated in Chapter 5.
161. In some books, load branch voltages are called phase voltages and load branch currents are called phase currents. To avoid confusion, we use the terms phase voltages and phase currents only in connection with three-phase sources.
162. Standard line voltages vary somewhat from one region to another and even within regions; for example, 110 Vrms, 117 Vrms, and 120 Vrms all are standard wall-outlet voltages.
163. Recall that the available source voltage is the open-circuit (Thévenin-equivalent) source voltage and the available source current is the short-circuit (Norton-equivalent) source current. Bear in mind that the available voltage (or current) is not necessarily the voltage across (or current entering) the input terminals of the circuit.
164. As opposed to arising as a model for an existing component or circuit.
165. Although the name seems to imply otherwise, gain is not necessarily greater than unity at any frequency.
166. The threshold of hearing depends upon frequency as well as incident power.
167. After Hendrik Bode (bōdee) (1905-1982), an engineer who showed how to use such plots to determine whether a system is stable and if not, how to make it so. Bode contributed much to our understanding of feedback systems.
168. Gain also could be expressed in terms of peak amplitudes; however, it usually is easier to obtain accurate values for peak-to-peak amplitudes from an oscilloscope display than it is to obtain accurate readings of peak amplitudes - especially if a dc component also is present.
169. A harmonic series is a sum of sinusoids where the frequency of every component is an integer multiple of a fundamental frequency.
170. A tone whose frequency is an integer multiple of the frequency of another tone is called a harmonic of the latter. In music, multiples of a basic note (the fundamental frequency) are sometimes called overtones, where the first overtone is the second harmonic.
171. Peter Gustave Lejeune Dirichlet (1805-1859), a German mathematician.
172. In this chapter, we use xrms to denote the rms amplitude of a signal x(t) because we use X to denote the Fourier coefficients of the signal.
173. In the table, symmetric means even symmetry about the vertical axis.
174. See G. Moretti, Functions of a Complex Variable, Prentice-Hall, 1964, pp 334-337.
175. After Josiah Willard Gibbs (1839-1903), a mathematical physicist who first explained the phenomenon.
176. However, if the source is capacitance coupled to the amplifier, the required dc bias-current compensation drastically reduces the input impedance.
177. Another is slew rate, discussed in Section 17.5.
178. We assume here that the bandwidth of the feedback amplifier exceeds that of the op amp alone, which is almost always the case (except for a follower). Because the gain-bandwidth product for a feedback amplifier equals that for the op amp alone, and because the dc gain of a stable feedback amplifier is smaller than that of the op amp, the bandwidth of the amplifier must be larger than that of the op amp.
179. The output swing of an op amp is defined as if the supply is symmetric, whether or not that is actually the case.
180. It also is possible to use a circuit called an automatic gain control (AGC), which keeps the rms amplitude of an input within specified bounds. You might learn about AGCs in a subsequent course.
181. You might find it helpful to review Section 8.10 of Chapter 8 before proceeding
182. See Section 8.10.7 in Chapter 8.
183. The output stage of an op amp is typically a push-pull configuration of two transistors, where one transistor amplifies the positive parts of an applied voltage and the other amplifies the negative parts. There is actually small crossover voltage range near zero where there is current through both supplies. However, this current is usually quite small, relative to a typical load current.
184. Actually, this is only the power delivered by the supply to the output stage. A typical op amp is a three-stage amplifier, and even if no input is applied (even if i0 = 0), some dc power is required to keep the op amp in an active (ready) state.
185. Direct coupling is used almost exclusively in integrated circuits.
186. The 14.14 mV source voltage shown is peak. The meter reading is rms.
187. Virtually all manufacturers of op amps have extensive web sites describing their offerings.
188. The integral in (18.2) is an example of a line integral, so called because the path of integration is along a line defined by s = c that is not necessarily coincident with one of the axes. You need not be alarmed by this integral because we never need to use it.
189. In this book, signal means current or voltage.
190. You will see this proved if you take a course in complex variables.
191. In physical problems, we are free to choose a time origin. We could generalize by treating right-sided functions that equal zero for t<0, but that would clutter the development without providing any useful increase in generality.
192. This requirement is somewhat stronger than necessary, but is easily tested and completely satisfactory in all practical applications.
193. Please accept on faith that the result is consistent with that of a rigorous development, such as provided by the theory of generalized functions (or distribution theory).
194. After Paul Dirac (1902-1984), a British physicist credited with introducing the delta function.
195. In general, the dimension of a delta function is the reciprocal of the dimension of the argument.
196. The left sides of the referenced relations can be defined in a self-consistent manner, but we do not need those more general relations in this book.
197. Entries 11-13 are useful mainly for obtaining inverse Laplace transforms.
198. 11Note that A is associated with p, which has a positive imaginary part and A* is associated with p*.
199. Recorded data or signals can be advanced relative to the time at which they were recorded, but that is not a true advance because it must be preceded by a delay (the recording time) sufficient to allow the advance.
200. After the British physicist Oliver Heaviside (1850-1925), who contributed much to electrical engineering.
201. The ideal differentiator described in Chapter 8 is rarely (if ever) used in practice.
202. In truth, frequency-domain analysis can be generalized through the Fourier transformation, and resulting properties and methods allow analysis of virtually any linear circuit and transformable excitation, not just sinusoids. You will learn about the Fourier transformation if you take a subsequent course in signal processing or communication systems, where it is favored over the Laplace transformation.
203. There are several definitions of stability. The one given here is called the BIBO (bounded-input-bounded output) definition and is an appropriate definition for linear circuit models. Note that no physical circuit can actually produce an unbounded output. Mathematical studies of stability necessarily deal with circuit models and thus with mathematical models of currents and voltages.
204. One can cook up excitations whose zeros cancel (mathematically) one or more poles of a transfer function, but in the real world, such cancellation is imperfect and all unforced terms are present, regardless of the form of the excitation.
205. Invented in 1955 by R.P. Sallen and E.L. Key of MIT Lincoln Laboratory.
206. Switched-capacitor resistors are discussed in Chapter 8.
207. The resistors denoted by 10R provide dc feedback for the integrators.
208. Look up Butterworth filter in Wikipedia for more complete descriptions of these filter types.
209. Horowitz, Paul and Winfield Hill, The Art of Electronics, Cambridge University Press, New York, 1989, pp 273-276.
210. Adapted from Table 5.2 in Horowitz and Hill (ibid).
211. Ability to achieve frequency-independent delay in the passband is one of the more important justifications for using certain kinds of digital filters.